NIKE Data Scientist Interview Questions and Answers

Case Study was related to customer propensity to buy.

Q1. Linear Regression Assumptions
Ans. 
Linear regression assumptions include linearity, independence, homoscedasticity, and normality.
• Assumption of linearity: The relationship between the independent and dependent variables is linear.

• Assumption of independence: The residuals are independent of each other.

• Assumption of homoscedasticity: The variance of the residuals is constant across all levels of the independent variables.

• Assumption of normality: The resid...read more
Answered by AI
Add your answer
Q2. ML Algorithms and Evaluation Metrics.
Add your answer

Q1. What is VIF(variance inflation factor)
Ans. 
VIF is a measure of multicollinearity in regression analysis, indicating how much the variance of an estimated regression coefficient is increased due to collinearity.
• VIF values greater than 10 indicate high multicollinearity

• VIF is calculated for each predictor variable in a regression model

• VIF is calculated as 1 / (1 - R^2) where R^2 is the coefficient of determination from regressing a predictor on all other predicto
Answered by AI
Add your answer

Q1. Why do you want to join?
Ans. 
I am impressed by your company's innovative projects and collaborative work culture.
• I admire the company's commitment to cutting-edge technology and data-driven solutions.

• I am excited about the opportunity to work with a talented team of data scientists and researchers.

• Your company's reputation for fostering a collaborative and inclusive work environment is appealing to me.
Answered by AI
Add your answer

Q1. How to reduce model inference latency
Ans. 
To reduce model inference latency, optimize model architecture, use efficient algorithms, batch processing, and deploy on high-performance hardware.
• Optimize model architecture by reducing complexity and removing unnecessary layers

• Use efficient algorithms like XGBoost or LightGBM for faster predictions

• Implement batch processing to make predictions in bulk rather than one at a time

• Deploy the model on high-performance har
Answered by AI
Add your answer
Q2. Different sql joins and their difference
Ans. 
SQL joins are used to combine rows from two or more tables based on a related column between them.
• INNER JOIN: Returns rows when there is at least one match in both tables.

• LEFT JOIN: Returns all rows from the left table and the matched rows from the right table.

• RIGHT JOIN: Returns all rows from the right table and the matched rows from the left table.

• FULL JOIN: Returns rows when there is a match in one of the tables.

• SEL
Answered by AI
Add your answer

Round duration - 60 Minutes
Round difficulty - Medium

I was asked two questions in this round . More emphasis was given on the theoretical aspect of the subject in this round .

Q1. How can you tune the hyperparameters of the XGBoost algorithm?
Ans. 
Hyperparameters of XGBoost can be tuned using techniques like grid search, random search, and Bayesian optimization.
• Use grid search to exhaustively search through a specified parameter grid

• Utilize random search to randomly sample hyperparameters from a specified distribution

• Apply Bayesian optimization to sequentially choose hyperparameters based on the outcomes of previous iterations
Answered by AI
View 1 more answer
Q2. Can you explain the hyperparameters in the XGBoost algorithm?
Ans. 
Hyperparameters in XGBoost algorithm control the behavior of the model during training.
• Hyperparameters include parameters like learning rate, max depth, number of trees, etc.

• They are set before the training process and can greatly impact the model's performance.

• Example: 'learning_rate': 0.1, 'max_depth': 5, 'n_estimators': 100
Answered by AI
View 1 more answer

Round duration - 50 Minutes
Round difficulty - Medium

This round basically tested some fundamental concepts related to Machine Learning and proper ways to implement a model.

Q1. What is the difference between Ridge and LASSO regression?
Ans. 
Ridge and LASSO regression are both regularization techniques used in linear regression to prevent overfitting by adding penalty terms to the cost function.
• Ridge regression adds a penalty term equivalent to the square of the magnitude of coefficients (L2 regularization).

• LASSO regression adds a penalty term equivalent to the absolute value of the magnitude of coefficients (L1 regularization).

• Ridge regression tends to sh...read more
Answered by AI
Add your answer

Round duration - 50 Minutes
Round difficulty - Medium

This round was based on some basic concepts revolving around Deep Learning .

Q1. What are outlier values and how do you treat them?
Ans. 
Outlier values are data points that significantly differ from the rest of the data, potentially affecting the analysis.
• Outliers can be identified using statistical methods like Z-score or IQR.

• Treatment options include removing outliers, transforming the data, or using robust statistical methods.

• Example: In a dataset of salaries, a value much higher or lower than the rest may be considered an outlier.
Answered by AI
Add your answer

Round duration - 30 Minutes
Round difficulty - Easy

This is a cultural fitment testing round .HR was very frank and asked standard questions. Then we discussed about my role.

Q1. What do you know about the company?
Add your answer
Q2. Why should we hire you?
Add your answer

Q1. All the questiones were asked around CV. Mostly problems related to ML, DL, NLP, mathematics behind the algorithms, case studies, alternate solutions of popular use cases etc.
Add your answer

Q1. Same as round 1 but this round involved a lot of mathematical functions and derivations of several aspects of ML and DL. Also a lot of case studies were involved
Add your answer

Had to share my screen and they gave live problems to test my knowledge in python

Q1. Call with hiring manager mostly on my CV and a lot of case studies.
Add your answer

Q1. Typical HR round questions
Add your answer

Q1. Why do you want to leave your current job?
Add your answer

SQL coding question. Medium level

Explain my project and then case study regarding launching new apps

Q1. Linear regression
Add your answer
Q2. Probability related questions
Add your answer
Q3. Sampling and AB testing
Add your answer

Q1. Backpropagation in neural network
Ans. 
Backpropagation is a method used to train neural networks by adjusting the weights based on the error in the output.
• Backpropagation involves calculating the gradient of the loss function with respect to the weights of the network.

• The gradient is then used to update the weights in the opposite direction to minimize the error.

• This process is repeated iteratively until the network converges to a solution.

• Backpropagation i...read more
Answered by AI
Add your answer
Q2. Clustering (k-means, DB scan)
Add your answer

1 question on array (sorting related), 1 question on string (hard problem)

Q1. Behavioral questions
Add your answer

3 Leet code mediums in 30 mins.

Q1. 5 ML questions in 10 mins
Add your answer
Q2. 5 Stats question in 10 mins
Add your answer
Q3. 3 LC mediums in 30 minutes
Ans. 
LC mediums refer to LeetCode mediums, which are medium difficulty coding problems on the LeetCode platform.
• LC mediums are coding problems with medium difficulty level on LeetCode platform.

• Solving 3 LC mediums in 30 minutes requires good problem-solving skills and efficient coding techniques.

• Examples of LC mediums include 'Longest Substring Without Repeating Characters' and 'Container With Most Water'.
Answered by AI
Add your answer